| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-aider-polyglot | OpenThinkerAgent-32B | #6 | Accuracy: 32.4 |
| code-generation-on-terminal-bench-2-0-terminus-2 | OpenThinkerAgent-32B (Terminus-2) | #10 | Accuracy: 26.2 |
| code-generation-on-swe-bench-verified | OpenThinkerAgent-32B (Terminus-2) | #73 | Accuracy: 54.0 |